PSA for time nuts: The #AppArmor profiles in #Debian Trixie for #gpsd and #chrony are broken if you try to use SOCK rather than SHM, for maximum #PPS accuracy. Bit me on #Proxmox 9.1. The #NTP SHM segment has to be polled by definition; SOCK is event-driven, so you only get #UNIX socket latency on pulses. So just put them in complain. Not an issue for me yet. The #SiRFStar IV #USB GPS I scored yesterday doesn't appear to do PPS sadly. Here's hoping, hoping, I stashed my GPSDO in safekeeping!

It appears that the #Intel #I219V e1000e driver for #Linux now supports #IEEE #1588 #PTPv2 precision timing with hardware #timestamping on the #ThinkPad P14S. I'm still qualifying it against an #I226V based master clock on my #LAN across an #FS #S3900 (non-retiming, no boundary clock) in P2P mode. The master is hosted on #Proxmox 9.1, and has a PCIe serial card ready for a GPSDO. There is also a #SiRFStar IV based USB GPS for it to fix to #utc but sadly it doesn't emit# PPS pulses on DCD.

I now have a working #IEEE #1588 hardware-corrected #time sync setup using an #Intel #I226V card. No #Arista switch required. FreeBSD #ptpd slave sees #linuxptp master. Yet to test with the #NVidia #Mellanox #ConnectX-4 LX clients in #FreeBSD though. #BeagleBone Black support appears broken. Awaiting a #SiRFStar IV USB #GPS with alleged #PPS capability, and yet to recover my #GPSDO unit from my aunt's attic in Fife.